Lamellipodial Versus Filopodial Mode of the Actin
Nanomachinery: Pivotal Role of the Filament Barbed End
The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.
Citation
Mejillano, Marisan R., Shin-ichiro Kojima, Derek Anthony
Applewhite, Frank B. Gertler, Tatyana M. Svitkina, and Gary G.
Borisy. “Lamellipodial Versus Filopodial Mode of the Actin
Nanomachinery.” Cell 118, no. 3 (August 2004): 363-373.
Copyright © 2004 Cell Press
As Published
http://dx.doi.org/10.1016/j.cell.2004.07.019
Publisher
Elsevier
Version
Final published version
Accessed
Wed May 25 19:02:24 EDT 2016
Citable Link
http://hdl.handle.net/1721.1/83498
Terms of Use
Article is made available in accordance with the publisher's policy
and may be subject to US copyright law. Please refer to the
publisher's site for terms of use.
Detailed Terms
Cell, Vol. 118, 363–373, August 6, 2004, Copyright 2004 by Cell Press
Lamellipodial Versus Filopodial Mode
of the Actin Nanomachinery:
Pivotal Role of the Filament Barbed End
Marisan R. Mejillano,1 Shin-ichiro Kojima,1
Derek Anthony Applewhite,1 Frank B. Gertler,2
Tatyana M. Svitkina,3 and Gary G. Borisy1,*
1
Department of Cell and Molecular Biology
Feinberg School of Medicine
Northwestern University
303 E. Chicago Avenue
Chicago, Illinois 60611
2
Department of Biology
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
3
Department of Biology
University of Pennsylvania
Philadelphia, Pennsylvania 19104
Summary
Understanding how a particular cell type expresses
the lamellipodial or filopodial form of the actin machinery is essential to understanding a cell’s functional
interactions. To determine how a cell “chooses” among
these alternative modes of “molecular hardware,” we
tested the role of key proteins that affect actin filament
barbed ends. Depletion of capping protein (CP) by
short hairpin RNA (shRNA) caused loss of lamellipodia
and explosive formation of filopodia. The knockdown
phenotype was rescued by a CP mutant refractory to
shRNA, but not by another barbed-end capper, gelsolin, demonstrating that the phenotype was specific
for CP. In Ena/VASP deficient cells, CP depletion resulted in ruffling instead of filopodia. We propose a
model for selection of lamellipodial versus filopodial
organization in which CP is a negative regulator of
filopodia formation and Ena/VASP has recruiting/activating functions downstream of actin filament elongation in addition to its previously suggested anticapping
and antibranching activities.
Introduction
Two alternate forms of actin machinery coexist at the
leading edge of most motile cells: lamellipodia which
seem designed for persistent protrusion over a surface,
and filopodia which appear to perform sensory and exploratory functions to steer cells depending on cues
from the environment. Although lamellipodia have been
investigated since the pioneering work of Abercrombie
(1970), filopodia have emerged in recent years as important in a broad range of cell biological processes, including epithelial sheet closure in development, wound healing, neuronal path finding, immune cell function, cell
invasion, and metastasis of cancer cells (Jacinto and
Wolpert, 2001; Rorth, 2003).
Although most cells in culture express both lamellipodia and filopodia, some cells, such as keratocytes
or neutrophils, express lamellipodia almost exclusively;
*Correspondence: g-borisy@northwestern.edu
others, such as dendritic cells or the growth cones of
neurons, are dominated by filopodia. Thus, understanding how a particular cell type expresses one or the other
form of the actin machinery is essential to understanding
a cell’s functional interactions. An emerging general issue in cell biology is how the nanomachinery of a cell
self-organizes. In the context of the actin machinery for
protrusive motility, this issue may be expressed as how
the cell “chooses” between lamellipodial and filopodial
modes of the “molecular hardware.”
Two models have been proposed to account for the
formation of actin-based protrusions in crawling cells:
the dendritic nucleation model for lamellipodia (Mullins
et al., 1998; Pollard and Borisy, 2003; Svitkina and Borisy, 1999) and the convergent elongation model for filopodia (Svitkina et al., 2003). In lamellipodia, branched
nucleation of actin is driven by activation of the Arp2/3
complex, followed by filament elongation and barbedend capping. The necessity for barbed-end capping in
lamellipodial protrusion is counterintuitive since it terminates filament elongation, which is the driving force for
motility (Pollard and Borisy, 2003). However, capping
is necessary to keep filaments short and their number
constant (Borisy and Svitkina, 2000; Carlier, 1998; Cooper and Schafer, 2000). In filopodia, both branching and
capping need to be prevented to allow continuous elongation of parallel filaments at their tip (Katoh et al., 1999;
Mallavarapu and Mitchison, 1999). In the convergentelongation model (Svitkina et al., 2003; Vignjevic et al.,
2003), filopodia are initiated from a dendritic network
by action of a tip complex that protects filaments against
capping allowing for their elongation coordinated with
a bundling activity. Thus, these models focus on the
elongation status of the actin filament barbed end as a
crucial determinant of whether lamellipodial or filopodial
protrusion results.
The elongation status of the barbed end may be considered as the net result of an interplay between capping
and anti-capping activities. Capping protein (CP), an ␣/␤
heterodimer (Mr ␣ ⵑ36 kDa; ␤ ⵑ32 kDa) (Schafer et al.,
1994) is considered the major barbed-end terminator
during cell motility because it (a) has high affinity (Kd ⫽
0.1–1 nM) for actin filament barbed ends in vitro, (b) is
ubiquitous in eukaryotes (Cooper et al., 1999), (c) is
enriched in lamellipodia (Schafer et al., 1998), and (d)
supports in vitro reconstituted bacterial comet tail motility (Loisel et al., 1999). Furthermore, loss of function or
null mutations in genetic systems such as yeast (Amatruda et al., 1990, 1992; Sizonenko et al., 1996), Drosophila (Hopmann et al., 1996), and Dictyostelium (Hug et al.,
1995) resulted in aberrant actin architecture. The Ena/
VASP family of proteins have been shown to promote
formation of longer and less branched actin filaments
at the leading edge in vivo and to be present at the tips
of filopodia (Bear et al., 2002). Thus, Ena/VASP have
been considered to have anticapping activity and are
therefore candidate effectors controlling barbed ends.
Models of lamellipodial and filopodial formation predict barbed-end capping lies at the fork in the road of
these two pathways. Consequently, we tested whether
Cell
364
proteins that affect actin filament barbed ends could
determine which form of the actin machinery was selected. Experiments were carried out in highly motile
cells expressing both protrusive structures to facilitate
observation of alternate phenotypes. Specifically, we
tested whether barbed-end capping by CP and anticapping by Ena/VASP proteins are key elements in the
mechanism of the lamellipodial/filopodial transition. We
report that CP depletion resulted in inhibition of lamellipodia and explosive formation of filopodia. Contrary to
expectations based on only an anticapping role for Ena/
VASP, CP depletion in the absence of Ena/VASP resulted in ruffling, not filopodia. We propose a model
for regulation of the lamellipodial/filopodial transition in
which CP is a negative regulator of filopodia formation
and Ena/VASP has functions downstream of actin filament elongation in addition to its previously suggested
anticapping activity (Bear et al., 2002).
Results
Strategy of CP Depletion and Rescue
Targeted depletion of CP by RNA interference was used
here as an approach to investigate CP function in cultured cells. For this purpose, we prepared a hairpin
siRNA expression vector with a GFP marker (Kojima et
al., 2004) for the added advantage of easier detection
and sorting of expressing cells. Vertebrates have three
␣ isoforms, each encoded by a separate gene, and three
␤ isoforms produced from a single gene by alternative
splicing. ␤1 is found mostly in muscle (at the Z-line), ␤2
is widespread in nonmuscle cells (Schafer et al., 1994),
and ␤3 is present exclusively in male germ cells (von
Bulow et al., 1997). Two different 19-nucleotide targeting
sequences (T1, T2; Figure 1A) were selected that were
common to mouse and rat CP ␤1 and ␤2 mRNA-coding
regions while not having significant homology to any
other known genes in the mouse or rat database as
determined by BLAST search (NCBI). Since functional
CP exists only as a heterodimer, silencing of the
␤-subunit was predicted to lead to functional loss of
capping activity (Amatruda et al., 1992; Schafer et al.,
1992).
Controls for the specificity of silencing involved two
approaches: expression of siRNAs with mismatching
nucleotides and rescue of the knockdown phenotype
by a CP gene that was refractory to silencing. siRNAs
are known to be highly specific and modification of only
one or two base pairs in the target sequence is enough
to abrogate silencing. Therefore, we designed hairpin
constructs, T1* and T2*, identical to T1 and T2, respectively, except for two base pair mismatches in the target
sequence (Figure 1A). CP ␤2 cDNAs refractory to silencing were designed by introducing mutations in the third
nucleotide position of two codons (identical to those in
T1* and T2*) that did not affect the amino acid sequence.
These silent mutations would produce mutant mRNAs
refractory to RNAi, thus permitting expression of wildtype protein that could support rescue. Empty pG-Super
vector that expresses just the soluble GFP marker was
used as a control for transfection and GFP expression.
Figure 1. Expression of Hairpin siRNA Results in CP Depletion in
B16F1 Cells
(A) Target sequences, T1 and T2, selected for mouse and rat ␤1
and ␤2 mRNA. Control mismatch target sequences, T1* and T2*,
contain two nucleotide changes (red). Nucleotide numbers from
start codon of mouse ␤2 subunit are indicated.
(B) Immunoblots and densitometry results of lysates prepared from
FACS-purified cell populations five days after transfection of shRNA
constructs. The same blot was first probed with CP␣ antibody (␣),
then with the pan CP␤ (R22) antibody (␤); ␣-tubulin served as loading
control. Bar chart shows relative protein levels from densitometry
analysis of the immunoblots shown above. Protein amount from
lysates expressing pG-Super empty vector was assigned 100%.
CP␣, white bars; CP␤ black bars. Knockdown of CP ␤ subunit also
resulted in loss of ␣ subunit.
CP Depletion in B16F1 Cells
Detailed analysis of the CP knockdown phenotype was
performed using the mouse melanoma B16F1 cell line
because these cells are highly motile and express lamellipodia and filopodia. Cells expressing T1- or T2 shRNA
were surprisingly viable and could be grown in culture
for more than a week. GFP-expressing cells were collected by fluorescence-activated cell sorting (FACS) one
day after transfection (⬎99% GFP positive in the sorted
population) and analyzed 3–5 days later. Microscopic
inspection of the sorted cell population showed that
⬎90% of cells were still GFP positive after five days.
Immunoblotting confirmed silencing at the protein level
(Figure 1B) with the average reduction in CP being 95%
for T1 and 90% for T2. Control transfections with T1*
and T2* hairpin constructs did not result in decreased
CP amounts compared to transfection with GFP alone,
demonstrating that knockdown achieved by T1 or T2
siRNA was specific. Consistent with earlier studies that
CP is stable only as a heterodimer (Amatruda et al.,
1992; Schafer et al., 1992), knockdown of the ␤-subunit
Mode of Actin Protrusive Machinery
365
Figure 2. Phenotype
B16F1 Cells
of
CP
Knockdown
(A) Distribution of actin revealed by phalloidin
staining (left column) and CP revealed by R26
antibody (middle column) in fixed cells expressing pG-Super-T2* (upper row) or pGSuper-T2 (lower row) for 5 days. Right column
shows merged images. Boxed regions are enlarged in insets. T2*-expressing cell exhibits
normal morphology with CP localized at extreme leading edge of lamellipodia and excluded from filopodia (insets). T2-expressing
cell displays numerous filopodia, reduction
of lamellipodia, and no CP staining at the cell
periphery. Residual diffuse fluorescence in
thick cell regions is attributed to nonspecific
binding of secondary antibody. Scale bar is
equal to 10 ␮m.
(B) Phase contrast (upper row) and GFP fluorescence (lower row) images of live control
(T1*) and knockdown (T1 and T2) cells four
days after transfection with respective pGSuper constructs demonstrate increase in lamella density (phase) and thickness (GFP)
after CP knockdown. Scale bar is equal to
10 ␮m.
(C and D) Quantification of lamellipodia reduction (C) or filopodia enrichment (D) after
CP knockdown. Bars represent average %
of cell perimeter occupied by lamellipodial
network (C) or average number of filopodia
(D) in phalloidin-stained cells transfected with
indicated shRNA constructs (n ⫽ 16–20 cells).
CP induced changes were significant ( p ⬍
0.0001) for both lamellipodia and filopodia as
determined by Student’s t test.
(E) Stars formed at ventral surface of knockdown cells detected by phalloidin staining
(left) and by phase contrast microscopy
(right). Insets show enlarged examples. Scale
bar is equal to 5 ␮m.
also resulted in comparable depletion of the ␣-subunit
(Figure 1B).
CP depletion was assayed at the cellular level by immunostaining with CP antibody. Phalloidin was used to
visualize actin in the same cells. CP in control cells
(Figure 2A, upper row) was enriched at the extreme
leading edge of lamellipodia and was excluded from
filopodia, as demonstrated previously (Svitkina et al.,
2003). In knockdown cells (Figure 2A, lower row), CP
labeling was abolished at the leading edge, consistent
with reduction in CP protein levels shown by immunoblotting.
Attenuation of Lamellipodia Formation
by CP Knockdown in B16F1 Cells
Control B16F1 cells (nontransfected cells, cells expressing GFP alone or T1* and T2* constructs, Figure 2A,
upper row), growing on laminin-coated glass protruded
large thin lamella with broad actin-rich lamellipodia at
their leading edge, and a variable number of filopodia
partially or completely embedded in the lamellipodia.
However, in knockdown cells, lamellipodia were significantly attenuated. CP depletion caused cells to lose
their characteristic flat, thin morphology with a convex
leading edge and the cell perimeter lost its smooth outline. The whole leading lamella extending from the cell
edge toward the cell body appeared much denser in
phase contrast images (Figure 2B, upper row); in fluorescence images, they had higher intensity of soluble GFP
because of increased thickness (Figure 2B, lower row).
Phalloidin staining revealed a significant decrease in
the lamellipodial actin network and an increase in actin
staining in deeper lamellar regions (Figure 2A, lower row)
compared to cells expressing GFP alone or mismatched
hairpin constructs (Figure 2A, upper row). Morphometric
analysis showed a 3–4-fold decrease in the percentage
of lamellipodial perimeter associated with diffuse actin
network (Figure 2C), whereas overall cell shape and
spread area were not dramatically changed.
Attenuation of lamellipodia was confirmed by dimin-
Cell
366
Figure 3. Displacement of the Arp2/3 Complex in Knockdown B16F1 Cells
(A) Distribution of Arp2/3 complex revealed
by p16-Arc antibody (middle row) and actin
by phalloidin staining (top row) in fixed cells
expressing pG-Super-T2* (left column) or pGSuper-T2 (middle column) for five days. Bottom row shows merged images. Boxed regions are enlarged in insets. T2*-expressing
cell exhibits normal morphology with Arp2/3
complex localized throughout the lamellipodium and excluded from filopodia (insets). T2expressing cell displays numerous filopodia
and is almost devoid of lamellipodia, with
Arp2/3 labeling absent from periphery but
with some diffuse staining observed in the
cytoplasm. Scale bar is equal to 10 ␮m. Right
column shows Arp2/3 and actin distribution in
stars, with Arp2/3 localized only at the center.
Scale bar is equal to 5 ␮m.
(B) Immunoblots performed five days after
shRNA transfection with indicated constructs
and probed with either p16 (left) or Arp3
(right) antibody indicate no decrease in Arp2/3
complex protein levels after CP knockdown.
␣-Tubulin antibody probing was used as
loading control.
ished Arp2/3 staining at the leading edge (Figure 3A).
In control cells, Arp2/3 complex was enriched in lamellipodia and substantially colocalized with lamellipodial
actin but was significantly diminished at the perimeter
of knockdown cells. However, protein levels of two different Arp2/3 complex subunits (p16 and Arp3) were
unchanged as shown by Western blotting (Figure 3B),
indicating that lamellipodial attenuation was not due to a
reduction in overall cellular levels of the Arp2/3 complex.
Thus, lamellipodial attenuation by CP depletion was accompanied by a redistribution of the Arp2/3 complex
so that it was no longer enriched at the cell perimeter.
Enhanced Filopodia Formation
in CP Knockdown B16F1 Cells
Concomitantly with lamellipodia attenuation, CP depletion caused explosive formation of filopodia (Figure 2A,
lower row). Increased filopodia were apparent as early
as three days after siRNA-plasmid transfection and the
increase was confirmed by quantitative analysis (Figure
2D). Filopodia formation in knockdown cells occurred
not only at the cell edges, but also on the dorsal cell
surface. Since it was feasible to count only those filopodia originating from cell edges, the real magnitude of
enhanced filopodia formation may be greater than estimated.
In addition to peripheral and dorsal filopodia, star-
like structures which could be identified both in phasecontrast images and after phalloidin staining (Figure 2E)
were present at the ventral cell surface of 60 ⫾ 15%
(100 cells, n ⫽ 3 experiments) of knockdown cells. These
“stars” were found in lamellar and perinuclear regions.
They consisted of actin-containing spikes radiating from
a center brightly stained with phalloidin.
A hallmark of genuine filopodia is enrichment of fascin,
an actin bundling protein, along the length of filopodia
(Kureishy et al., 2002), and VASP, a barbed-end binding
protein (Bear et al., 2002), at their tips (Lanier et al.,
1999; Rottner et al., 1999). Cellular levels of both proteins
remained unchanged after CP knockdown in B16F1 cells
as determined by immunoblotting (data not shown). By
immunofluorescence, all lateral, dorsal, or ventral spikes
showed fascin staining along their lengths (Figure 4A,
upper row) and VASP staining at their tips (Figure 4B,
upper row). A lamellipodial marker, Arp2/3 complex is
usually depleted from filopodial bundles (Svitkina and
Borisy, 1999). Consistent with this, we found that all
filopodia induced by CP knockdown lacked Arp2/3 complex (Figure 3A), including the ventral stars, which were
positive for Arp2/3 complex at their center but not along
their radii (Figure 3A, right column). The staining pattern
of ventral stars was similar to the filopodial-like bundles
reconstructed in vitro from beads coated with Arp2/3
activators (Vignjevic et al., 2003). Thus, molecular marker
Mode of Actin Protrusive Machinery
367
Figure 4. Filopodial Markers Have Normal Distribution in Filopodia
Induced by CP Knockdown.
Knockdown B16F1 cells expressing pG-Super-T1 construct for four
days show numerous lateral filopodia (left column), dorsal filopodia
(middle column) and ventral stars (right column).
(A) Antibody staining reveals fascin along the length of filopodia.
(B) VASP (upper row) and actin (middle row) costaining reveals VASP
at the filopodial tips. Merged images of VASP and actin are shown
in bottom row. Boxed regions are enlarged in insets. Scale bars are
equal to 10 ␮m (left and middle columns) and 5 ␮m (right column).
analyses confirmed that the spikes and stars induced
by CP depletion were similar to filopodia.
Electron microscopy was carried out to evaluate the
phenotype of CP knockdown in B16F1 cells in more
detail. Low magnification views showed that control
cells contained few filopodia that were restricted to the
cell periphery (Figure 5A), whereas CP knockdown cells
showed abundant filopodial-like protrusions around the
cell perimeter and on the dorsal surface (Figure 5C),
which significantly varied in thickness and appeared
emerging from the surrounding network. In addition, the
leading lamella of knockdown cells was filled with a
dense actin filament network, which contrasted with a
rather sparse lamella cytoskeleton in control cells and
suggested increased actin filament assembly away from
the leading edge after CP depletion. High resolution
analysis demonstrated that similar to normal filopodia
(Figure 5B), lateral (Figure 5D) and dorsal (inset in Figure
5C) protrusions in CP knockdown cells contained bundles of long unbranched actin filaments that extended
almost the entire length of the bundle. Thus, EM data
are also consistent with the idea that CP knockdowninduced genuine filopodia with proper structural organization.
Figure 5. Structural Organization of Filopodia in B16F1 Cells
(A and B) Control cells. Platinum replica EM shows few peripherally
located filopodia embedded into lamellipodial network. Deeper cytoplasm shows sparser filament network.
(C and D) CP knockdown cells. Abundant filopodia are apparent at
both leading edge and dorsal surface. Deeper cytoplasm shows
dense filament network. High magnification images of control (B)
and CP knockdown (D) filopodia show their similar structural organization. Enlarged image of boxed region in (C) shows that dorsal
filopodium contains a bundle of parallel filaments. Scale bars are
equal to 5 ␮m (A, C) and 1 ␮m (B, D).
Kinetic analysis was performed to determine the dynamics of induced filopodia and to evaluate the effect
of CP knockdown on the protrusion of the leading edge
(See Supplemental Movies S1 and S2 available at http://
www.cell.com/cgi/content/full/118/3/363/DC1). Time
lapse movies of knockdown cells demonstrated that
induced filopodia displayed normal dynamic features:
protrusive, retractile, and sweeping motility. Dorsal filo-
Cell
368
podia were not formed as a result of retrograde translocation but were formed de novo. When behavior of the
leading edge was analyzed using kymographs, control
cells displayed smooth and persistent protrusion for
long periods of time, whereas CP-depleted cells displayed erratic protrusion-withdrawal behavior with no
obvious coordination between adjacent regions of the
leading edge (Supplemental Figure S1, Supplemental
Movies S1 and S2 available on Cell website). As a result,
knockdown cells demonstrated significant decrease in
net protrusion. The average rate of leading edge advance in CP-depleted cells (0.32 ⫾ 0.36 ␮m/min) was
2.0⫻ lower as compared to their mismatch controls
(0.64 ⫾ 0.41 ␮m/min) (4 s frame rate; 5 min observation
period; 42 kymographs; 10 cells, p ⬍ 0.001).
Rescue of Knockdown Phenotype
We attempted to rescue CP depleted cells by expressing
CP refractory to the siRNA. Successful rescue would
not only provide a stringent test for the specificity and
reversibility of the knockdown phenotype; it would also
test whether CP tagged with fluorescent protein could
efficiently substitute for endogenous capping activity.
We constructed CP␤2 subunits containing silent mutations (T1*, T2*; Figure 1A). To monitor simultaneously
knockdown and rescue constructs in transfected cells,
CFP was substituted for GFP in the pG-Super T1 and
T2 knockdown constructs, and the rescue constructs,
T1*- and T2*-CP␤2, were expressed as YFP-fusion proteins. Thus, CFP was taken as the indicator of knockdown and YFP the indicator of rescue. Cells were first
transfected with knockdown constructs to allow phenotype development and on day 3 were transfected with
one of the rescue constructs.
The phenotype of expressing cells 2–3 days after
YFP-CP (T1* or T2*) transfection was analyzed by phase
contrast and fluorescence microscopy. Most cells expressing both CFP and YFP (90%; n ⫽ 101) exhibited
prominent flat, thin lamellipodia and distinct CP localization at the cell periphery, a phenotype characteristic of
control cells (Figure 6A, i ). These results also indicated
that adding back the ␤-subunit by transfection of YFPCP␤ restored CP ␣-subunit to normal cellular levels. In
the same population, cells expressing only CFP (100%;
n ⫽ 106) still showed the knockdown phenotype (Figure
6A, ii ), characterized by abundant filopodia and thick,
dense lamella. All YFP-only positive cells (n ⫽ 100) had
normal CP localization at the leading edge indistinguishable from wild-type (Figure 6A, iii ). Thus, expression of
the rescue construct induced phenotypic reversion in
the presence of inhibitory siRNA demonstrating specificity and lack of irreversible downstream effects of CP
depletion. Rescue also established that CP tagged with
derivatives of GFP could support normal capping
function.
Next, we tested whether rescue could be achieved
with another barbed-end capper, gelsolin, or whether
the phenotype was indeed specific for CP. We confirmed
that gelsolin is present in B16F1 melanoma cells by
immunoblotting (data not shown). Immunofluorescence
staining (Supplemental Figure S2 available on Cell website) showed that gelsolin is enriched in the lamellipodia,
consistent with earlier studies (Chou et al., 2002). However, compared to CP, which is present mostly at the
Figure 6. Rescue of CP Knockdown Phenotype in B16F1 Cells
(A) Rescue by CP refractory to siRNA. (i ) Cell expressing both knockdown (CFP-T1 hairpin) construct (left) and rescue construct (YFPCP(T1*)) refractory to siRNA (middle) shows wild-type phenotype
with YFP-CP(T1*) localizing to the extreme leading edge as in normal
cells. Phase image (right) shows flat and thin lamellipodial morphology indistinguishable from control cells. (ii ) Cell expressing CFP-T1
hairpin (left), but not YFP-CP(T1*) (middle), displays characteristic of
the knockdown phenotype with numerous spiky protrusions evident
from the phase contrast image (right) and thick lamella as revealed
by distribution of soluble CFP (left). (iii ) Cell expressing silent mutant
YFP-CP(T1*) (middle), but not CFP-T1 (left) shows wt distribution of
CP at extreme leading edge. Corresponding phase contrast image
(right) illustrates normal lamellipodial morphology. Scale bar is equal
to 10 ␮m.
(B) Gelsolin does not rescue CP knockdown phenotype. (i, ii ) Cells
expressing both CFP-T1 and YFP-gelsolin (i ), as well as cells expressing CFP-T1 alone (ii ) still show abundant filopodia and diminished lamellipodia (phase images), characteristic of CP knockdown
phenotype. (iii ) Cell expressing only YFP-gelsolin displays gelsolin
localization at the leading edge; corresponding phase image shows
control phenotype. Scale bar is equal to 10 ␮m.
extreme leading edge (see Figures 2A and 6A), gelsolin
distribution is broader, usually colocalizing with most of
the lamellipodial actin network. For rescue experiments
with gelsolin, we used the same strategy as for CP.
Mode of Actin Protrusive Machinery
369
Results showed that 100% of cells expressing both
CFP-T1 and YFP-gelsolin (n ⫽ 100) still showed the
CP knockdown phenotype characterized by excessive
filopodia and loss of lamellipodia, and that they were
indistinguishable from cells expressing only CFP-T1
(Figure 6B, i, ii ). Cells expressing only YFP-gelsolin exhibited gelsolin localization at the leading edge similar
to that revealed by immunostaining. In phase contrast
images, they displayed flat, thin lamellipodia, typical of
control cells (Figure 6B, iii ). These results show that
another barbed-end-capping protein, gelsolin, does not
functionally substitute for CP in these cells and suggests
that the ability to control the lamellipodial/filopodial transition requires properties specific to CP.
Generality of CP Knockdown Phenotype
To determine whether the phenotypic response of CP
depletion reflects a general principle, we performed CP
knockdown with shRNA in two other mammalian cell
types: a murine fibroblast cell line, NIH3T3, and a rat
fibroblast cell line, Rat2. Both cell types (Supplemental
Figures S3 and S4 available on Cell website) displayed
attenuation of lamellipodia and excessive formation of
filopodia after CP depletion. As in B16F1 cells (see Figure 2A), control Rat2 and NIH3T3 cells expressing T2*
construct showed that CP localized at the extreme leading edge and was excluded from filopodia; whereas in
cells expressing the T2 knockdown construct, CP staining was abolished at the cell edge. Phalloidin staining
and phase contrast microscopy revealed excessive filopodia and diminished (sometimes absent) lamellipodia
in knockdown cells, a phenotype essentially similar to
that of B16F1 cells depleted of CP.
CP Knockdown in Cells Lacking Ena/VASP
CP activity in vivo may be affected by Ena/VASP family
proteins that antagonize actin filament capping at the
lamellipodial leading edge (Bear et al., 2002). Therefore,
we performed CP depletion in MVD7 murine fibroblasts
lacking all endogenous Ena/VASP proteins (Bear et al.,
2000). The phenotype of these cells was evaluated
during spreading on laminin shortly (ⵑ30–45 min) after
replating, during which time cells displayed extensive
protrusive activity favorable for analysis. Control untransfected MVD7 cells, as well as the derivative cells
reexpressing Mena (MVD7 EM), displayed a smooth, circular lamellipodium under these conditions (Figures 7A
and 7B, upper rows). Less than 4% showed filopodia
and these were few in number. Less than 10% of untransfected cells showed any ruffling and this was restricted to a small fraction of the cell perimeter. However, after expression of the T1 knockdown construct
for five days, both cell types lost the smooth outline of
their cell perimeter, but they did so in strikingly different
ways (Figures 7A and 7B, lower rows). Most MVD7 EM
cells (73%), as expected, formed excessive filopodia
(44.0 ⫾ 2.0 per cell, n ⫽ 16 cells), whereas most MVD7
cells (60%) exhibited extensive ruffling all along the cell
perimeter, but not filopodia that were extremely rare
(4.0 ⫾ 0.2, n ⫽ 15; p ⬍ 0.001). These data confirm the
results with the other cell lines that CP depletion compromised lamellipodia formation. In addition, they
showed that CP depletion was not sufficient for filopodia
formation, but that Ena/VASP was required as well.
Figure 7. Phenotype of CP Depletion Depends on the Presence of
Ena/VASP Proteins
(A) Ena/VASP-deficient MVD7 cells spreading on laminin-coated
glass display the smooth outline of the cell perimeter in control
(upper image), but express numerous phase-dark ruffles after expressing CFP-T1 knockdown construct for five days (lower image)
as observed by phase contrast microscopy of living cells.
(B) Similar to its parent cell line, untransfected MVD7 reexpressing
GFP-Mena cells (MVD7 EM) (upper image), display smooth cell perimeter but show increased filopodial-like protrusions after expressing
CFP-T1 knockdown construct for five days. Enlarged insets show
GFP-Mena at extreme leading edge of control cell and at the tips
of filopodia-like protrusions in knockdown cell. Scale bar is equal
to 10 ␮m.
Discussion
CP-Dependent Control of the Lamellipodial/
Filopodial Transition
The main features of the CP depletion phenotype were
severe attenuation of lamellipodia and dramatic increase of filopodia. Lamellipodial attenuation was evi-
Cell
370
dent by the redistribution of Arp2/3 away from the cell
perimeter, by loss of the normal dendritic brush near
the leading edge and by appearance of a dense network
of filaments throughout the cytoplasm, suggestive of
uncontrolled polymerization. The loss of Arp2/3 at the
cell perimeter needs to be explained because the overall
cellular level of Arp2/3 was not affected by CP depletion
and lack of capping is not predicted to interfere with
branched nucleation. The simplest explanation is that
the filopodial machinery, which excludes Arp2/3 complex along the length of actin bundles, becomes so dominant at the cell perimeter as to force the redistribution
of Arp2/3. Another possibility is that the activities of CP
and Arp2/3 complex are tightly coordinated in cells to
produce a balanced dendritic network. Independent of
the explanation of the redistribution of Arp2/3, these
results indicate the positive requirement of CP for normal lamellipodial expression.
An even more striking result of CP knockdown was
the extensive formation of filopodia, not just at the leading edge but also at the dorsal and ventral surfaces of
cells. These results complement our previous findings
in vitro (Vignjevic et al., 2003) showing that plastic beads
coated with Arp2/3-activating proteins and placed into
CP-depleted brain cytoplasmic extracts were not able to
assemble comet tails, but assembled filament bundles in
star-like configuration strikingly similar to ventral stars
described here. The role of Arp2/3-activator coated
beads in cells may be played by Arp2/3-rich cytoplasmic
foci, which are abundant in B16F1 (Ballestrem et al.,
1998) and other cell types (Schafer et al., 1998). Taken
together, these results establish that CP is necessary
for formation of dendritic arrays and that decreased
CP concentrations favor transformation of the dendritic
network into filopodial bundles.
Although a number of barbed-end cappers exist in
cells (Schafer and Cooper, 1995), our data demonstrate
that the filopodial phenotype was produced specifically
by CP knockdown as it could be rescued by an RNAiresistant mutant of CP, but not by gelsolin, another
abundant ubiquitous barbed-end capper (Kwiatkowski,
1999). These findings agree with previous reports that
gelsolin null cells show normal filopodial protrusion and
display only a subtle phenotype with respect to their
protrusive behavior (Azuma et al., 1998). Although filopodia were reported to be more abundant in gelsolin
null neuronal growth cones as compared to wt, the increase was due to impaired retraction as opposed to
increased protrusion (Lu et al., 1997).
Based on our current findings and work from other
groups, we propose a model for the mechanism of regulation of lamellipodial versus filopodial protrusion by CP
(Figure 8). In this model, Arp2/3-dependent dendritic
nucleation of actin filaments is the starting point for
formation of both actin filament arrays, but then the
two pathways diverge depending on CP activity. We
propose that when CP activity in the cell is high (Figure
8, upper images), newly nucleated filaments elongate
for only a short period of time and become capped. To
maintain protrusion, capped filaments are replaced by
newly nucleated side branches produced by the Arp2/3
complex. This favors the formation of a dendritic lamellipodial network. In contrast, when activity of CP is low
(Figure 8, lower images), filaments can elongate persis-
tently and attain significant lengths. Other elements of
the convergent elongation mechanism work on these
preformed long filaments allowing their convergence
and subsequent bundling to form filopodia.
Control of Barbed-End Elongation
How is barbed-end elongation controlled? The mechanism of CP regulation is a long-standing issue that has
not yet been resolved. In vitro, capping activity of CP
can be inhibited by PIP2 micelles (Schafer et al., 1996),
but it is not known whether such regulation takes place
in vivo. No phosphorylation of CP has been reported
(Schafer et al., 1994). Recently, two proteins interacting
with CP have been identified. One of them, V-1, forms
a stable complex with CP and inhibits its binding to
actin filaments in vitro (Taoka et al., 2003). Another CPinteracting protein, CARMIL, was found in Dictyostelium
and Acanthamoeba (Jung et al., 2001; Remmert et al.,
2004), but it remains unknown whether it affects CP
activity in vitro or in vivo.
Regulation of barbed-end elongation may involve proteins that antagonize capping. One group of such proteins is the Ena/VASP family. Ena/VASP proteins when
targeted to the membrane promote formation of longer
and less branched actin filaments at the leading edge
and, conversely, depletion of Ena/VASP leads to shorter,
more branched filaments (Bear et al., 2002). VASP is also
enriched at filopodial tips (Lanier et al., 1999; Rottner et
al., 1999), and accumulates at sites on the lamellipodial
leading edge where filopodia emerge (Svitkina et al.,
2003). Deficiency of Ena/VASP in Dictyostelium (Han et
al., 2002) or neuronal cells (Lebrand et al., 2004) resulted
in inhibition of filopodia, whereas overexpression or
membrane targeting of VASP produced overabundant
filopodia in Dictyostelium, as well as increased filopodial
length and number in neurons. Ena/VASP proteins are
known to be regulated by PKA/PKG-dependent phosphorylation (Kwiatkowski et al., 2003) at multiple phosphorylation sites. The conserved N-terminal Ser residue
in vertebrate Ena/VASP proteins seems to be involved
in fibroblast motility (Loureiro et al., 2002). Moreover,
increased phosphorylation of this Ser in Mena in response to activation of PKA directly paralleled formation
of filopodia in neurons (Lebrand et al., 2004). Thus, capping antagonists in their active form may serve as positive regulators of actin filament elongation and, therefore, filopodia formation. Evidence that the balance
between capping and anticapping is critical to the
“choice” of actin organization that dominates in a specific cell type or site comes from genetic analysis in
Drosophila of microvilli and cleavage furrows (Grevengoed et al., 2003) and from immunoblot analysis of actin
binding proteins in cell cultures showing low levels of
CP and high levels of Ena/VASP in neurons as compared
to fibroblasts (Strasser et al., 2004).
In addition to Ena/VASP, another group of proteins
that antagonize capping has been identified recently,
namely, formins. In vitro, formins nucleate actin filaments and remain associated with their growing barbed
ends (Evangelista et al., 2003; Pruyne et al., 2002) while
protecting them from capping (Evangelista et al., 2003;
Harris et al., 2004; Moseley et al., 2004; Pruyne et al.,
2002; Zigmond et al., 2003). Therefore, regulation of
Mode of Actin Protrusive Machinery
371
Figure 8. Model for Mechanism of Regulation of Lamellipodial Versus Filopodial Protrusion by Filament Capping
Box on left illustrates starting point for both types of protrusion and possible regulation of the proposed key players. Smaller boxes illustrate
scenarios for lamellipodia (top) or filopodia (bottom) depending on capping activity: Top: when capping activity dominates either by activation
of CP or inhibition of capping antagonists such as Ena/VASP, filaments elongate for a brief period before becoming capped and, as a
consequence, these filaments are relatively short. To maintain protrusion, capped filaments are replaced by newly nucleated side branches
produced by Arp2/3 complex, favoring the formation of a dominantly branched, lamellipodial network. Bottom: low capping activity resulting
either from inhibition of CP or activation of proteins that promote anticapping and antibranching such as Ena/VASP, favor filament elongation
leading to long and unbranched filaments, which converge and subsequently become bundled by fascin. Filopodia formation is therefore
more predominant. CP inhibition may be mediated by PIP2, the CARMIL complex or the V-1 protein; Ena/VASP proteins may be activated by
PKA/PKG-dependent phosphorylation.
formins also has the potential to influence the balance
between filopodia and lamellipodia in cells.
Control of Filopodia Formation
Why does CP depletion lead to filopodia formation and
not just erratic protrusive behavior? According to the
convergent elongation model, filament elongation represents only the first step of transition from the dendritic
network to parallel bundles. Any missing or inactive element of the filopodial machinery downstream of filament
elongation, such as filament bundling, would prevent
filopodia formation and lead to uncontrolled polymerization. The fact that CP depletion led to explosive formation of filopodia indicates that the rest of the filopodiamaking machinery was present in the cells and ready
to go. This suggests that under normal conditions, control of the status of the actin filament barbed end at the
leading edge serves as a crucial determinant of whether
the lamellipodial or filopodial mode of the actin machinery is expressed.
Drosophila S2 cells treated with dsRNA against CP
displayed excessive ruffling, but not filopodia (Rogers
et al., 2003). Since filopodia were absent in S2 cells even
in normal culture, an intrinsic deficiency of the filopodial
machinery in this cell type is likely responsible for the
different phenotype. The observed ruffling can be interpreted as a manifestation of inefficient protrusion resulting from long filaments buckling under the membrane.
What are the elements of the filopodial machinery
necessary to complete filopodia formation after filament
elongation is allowed? In this study, we showed that an
Ena/VASP family member was essential for filopodia
formation. Without Ena/VASP but with normal levels of
CP, actin filaments were short and highly branched (Bear
et al., 2002). Depleting CP in Ena/VASP deficient cells
allowed for filament elongation. However, very few filopodia formed and cells switched to ruffling, similar to
Drosophila S2 cells. Cells reexpressing Ena/VASP formed
filopodia upon CP depletion. Thus, Ena/VASP proteins
are not only capping antagonists (Bear et al., 2002),
but also part of the filopodial machinery downstream of
elongation (this study). Downsteam functions of Ena/
VASP may involve establishing the functionality of the
filopodial tip complex (Svitkina et al., 2003). Alternatively, Ena/VASP may participate in recruiting fascin or
its activators, which are important for filament crosslinking, and therefore bundling (Vignjevic et al., 2003).
In summary, our analysis has demonstrated that key
proteins, CP and Ena/VASP, play a critical role at the
molecular “hardware” level of the actin protrusive machinery. They sit at a fork in the road of the dendriticnucleation and convergent-elongation pathways for
lamellipodia and filopodia formation, respectively. CP
serves as a negative regulator of filopodia and a positive
regulator of lamellipodia. Low levels of CP or high levels
of its antagonists determine the extent of prolonged
filament growth, which, if allowed, degrades the dendritic network, but allows for engagement of downstream machinery and filopodia formation. Ena/VASP
proteins play an important role not only in terms of
anticapping activity but also in the engagement of the
downstream filopodial machinery. If absent or deficient,
Cell
372
actin filament growth results in ruffling, not filopodia
formation. Although the full upstream pathway of regulation of cappers and anticappers remains to be elucidated, it is clear that such mechanisms exist and that
they are likely complex and intertwined. The salient point
of this study is that the elongation status of actin filament
barbed ends determines the expression of the lamellipodial or filopodial mode of actin protrusion, contingent
on the functionality of the downstream filopodial machinery.
Experimental Procedures
Expression Plasmids
pG-Super hairpin siRNA construct (Kojima et al., 2004) was based on
the pSuper expression vector (Brummelkamp et al., 2002). pEGFPCP␤2 was a gift from D. Schafer (University of Virginia). YFP-CP␤2
was obtained by transferring the CP␤2 cDNA into pEYFP-C1 (Clontech). Silent mutations in YFP-CP were introduced using the Quikchange site-directed mutagenesis kit II (Stratagene). pEYFP-gelsolin
was constructed based on a cDNA fragment of human cytoplasmic
gelsolin from pCMV-gelsolin, a gift from H.L. Yin (Univ. of Texas
Southwestern Medical Center), and pEYFP-C1 with their HindIIIBamHI sites. To remove the 5⬘-untranslated region, a XhoI-NotI part
in the 5⬘ region was replaced with a PCR-amplified fragment.
Antibodies
The following rabbit polyclonal antibodies were obtained as gifts:
CP␤2 C terminus (R26) and the pan-CP␤-subunit (R22) from D.
Schafer; Arp3 from T. Uruno (Holland American Red Cross Laboratory); and gelsolin from H.L. Yin. CP␣ antibody was generated in
rabbits against the peptide IESHQFQPKNFWNGRWRSEWK and
affinity purified. Arp2/3 complex p16-Arc subunit antibody (Vignjevic
et al., 2003) and VASP antibody (Bear et al., 2002) were described
previously. The following antibodies were from commercial sources:
fascin (DAKO Cytomation), ␣-tubulin (Sigma), secondary antibodies
(Sigma, Jackson Laboratories). All other reagents were from Sigma
unless indicated otherwise.
Cell Culture and Microscopy
B16F1 mouse melanoma cells were provided by C. Ballestrem (Weizmann Institute of Science) and cultured as described previously
(Ballestrem et al., 1998). NIH3T3 cells was purchased from ATCC
and cultured in DMEM supplemented with 10% FBS. Rat2, MVD7,
and MVD7 EM were cultured as described (Bear et al., 2000). Transient
expression of proteins and shRNA was performed with FuGENE6
(Roche) according to manufacturer’s recommendations. For live cell
imaging, cells were transferred into L-15 medium or phenol red-free
DMEM (Gibco) supplemented with 10% FBS.
Immunostaining was performed after cell extraction for 5 min with
1% Triton X-100 in PEM buffer (100 mM PIPES, [pH 6.9], 1 mM
MgCl2, 1 mM EGTA) containing 4% PEG, MW 40,000 (Serva) and
2 ␮M phalloidin, followed by fixation with 0.2% glutaraldehyde and
quenching with NaBH4. Texas red phalloidin (0.033 ␮M, Molecular
Probes) was used for actin staining. Fascin staining was performed
after methanol fixation of extracted cells. For VASP staining, cells
were extracted and fixed simultaneously using a mixture of 0.5%
Triton X-100 and 0.25% glutaraldehyde in PEM buffer. Primary antibody concentrations used were 5–50 ␮g/ml.
Light microscopy was performed using an inverted microscope
(Eclipse TE2000 or Diaphot 300, Nikon) equipped with Plan 100⫻
1.3 NA objective and a back-illuminated cooled CCD camera (model
CH250, Roper Scientific) driven by Metamorph imaging software
(Universal Imaging Corp). For live imaging, cells were kept on the
microscope stage at 37⬚C during observation.
Quantification of both lamellipodia and filopodia was done on
fixed cells stained with Texas red phalloidin. Only spread cells containing a prominent leading lamella were chosen. For quantification
of lamellipodia, Metamorph line function was used to trace and
measure the whole-cell perimeter and cell perimeter with adjacent
lamellipodial network. The fraction of cell perimeter occupied by
lamellipodial network was used as a parameter for quantification.
The number of filopodia per cell was determined by counting only
filopodia crossing the cell edge and having fluorescence intensity
1.2⫻ above background. Two researchers performed both measurements independently. Motility analysis was done using kymograph
function in Metamorph. Statistical significance was determined by
Student’s t test. Graphs and statistical analysis were done using
SigmaPlot software. Samples for platinum replica electron microscopy were processed as described (Svitkina and Borisy, 1998).
Immunoblotting and Densitometry
Cells were lysed in buffer containing 10 mM Tris, [pH 7.5], 150 mM
NaCl, 1% Triton X-100, 10% glycerol, and protease inhibitor tablet
(Roche). Protein concentration of the lysates was determined using
BCA reagent (Pierce). SDS-PAGE (4%–20% polyacrylamide) and
immunoblotting were according to standard protocols. Densitometry analysis was performed using NIH Image software.
Acknowledgments
We thank D. Schafer, T. Uruno, H. L. Yin, and C. Ballestrem for gifts
of reagents; O. Chaga for technical assistance with phase contrast
and electron microscopy; J. Peloquin for preparing CP␣ and p16
antibodies; A. Mongiu, O. Danciu, and G. Woodhead for help in
preparing constructs; M. Paniagua for assistance in FACS; and S.
Zigmond for critical reading of the manuscript. Supported by grants
NIH GM62431 to G.G.B. and NIH GM58801 to F.B.G.
Received: October 29, 2003
Revised: June 14, 2004
Accepted: June 18, 2004
Published: August 5, 2004
References
Abercrombie, M., Heaysman, J.E., and Pegrum, S.M. (1970). The
locomotion of fibroblasts in culture. II. “Ruffling. Exp. Cell Res.
60, 437–444.
Amatruda, J.F., Cannon, J.F., Tatchell, K., Hug, C., and Cooper, J.A.
(1990). Disruption of the actin cytoskeleton in yeast capping protein
mutants. Nature 344, 352–354.
Amatruda, J.F., Gattermeir, D.J., Karpova, T.S., and Cooper, J.A.
(1992). Effects of null mutations and overexpression of capping
protein on morphogenesis, actin distribution and polarized secretion
in yeast. J. Cell Biol. 119, 1151–1162.
Azuma, T., Witke, W., Stossel, T.P., Hartwig, J.H., and Kwiatkowski,
D.J. (1998). Gelsolin is a downstream effector of rac for fibroblast
motility. EMBO J. 17, 1362–1370.
Ballestrem, C., Wehrle-Haller, B., and Imhof, B.A. (1998). Actin dynamics in living mammalian cells. J. Cell Sci. 111, 1649–1658.
Bear, J.E., Loureiro, J.J., Libova, I., Fassler, R., Wehland, J., and
Gertler, F.B. (2000). Negative regulation of fibroblast motility by Ena/
VASP proteins. Cell 101, 717–728.
Bear, J.E., Svitkina, T.M., Krause, M., Schafer, D.A., Loureiro, J.J.,
Strasser, G.A., Maly, I.V., Chaga, O.Y., Cooper, J.A., Borisy, G.G.,
and Gertler, F.B. (2002). Antagonism between Ena/VASP proteins
and actin filament capping regulates fibroblast motility. Cell 109,
509–521.
Borisy, G.G., and Svitkina, T.M. (2000). Actin machinery: pushing
the envelope. Curr. Opin. Cell Biol. 12, 104–112.
Brummelkamp, T.R., Bernards, R., and Agami, R. (2002). A system
for stable expression of short interfering RNAs in mammalian cells.
Science 296, 550–553.
Carlier, M.F. (1998). Control of actin dynamics. Curr. Opin. Cell Biol.
10, 45–51.
Chou, J., Stolz, D.B., Burke, N.A., Watkins, S.C., and Wells, A. (2002).
Distribution of gelsolin and phosphoinositol 4,5-bisphosphate in lamellipodia during EGF-induced motility. Int. J. Biochem. Cell Biol.
34, 776–790.
Cooper, J.A., and Schafer, D.A. (2000). Control of actin assembly
and disassembly at filament ends. Curr. Opin. Cell Biol. 12, 97–103.
Cooper, J.A., Hart, M.C., Karpova, T.S., and Schafer, D.A. (1999).
Mode of Actin Protrusive Machinery
373
Guidebook to the cytoskeletal and motor proteins, 2nd ed. (Oxford:
Oxford University Press).
capping, and formation of branching networks of filaments. Proc.
Natl. Acad. Sci. USA 95, 6181–6186.
Evangelista, M., Zigmond, S., and Boone, C. (2003). Formins: signaling effectors for assembly and polarization of actin filaments. J. Cell
Sci. 116, 2603–2611.
Pollard, T.D., and Borisy, G.G. (2003). Cellular motility driven by
assembly and disassembly of actin filaments. Cell 112, 453–465.
Grevengoed, E.E., Fox, D.T., Gates, J., and Peifer, M. (2003). Balancing different types of actin polymerization at distinct sites: roles for
Abelson kinase and Enabled. J. Cell Biol. 163, 1267–1279.
Han, Y.H., Chung, C.Y., Wessels, D., Stephens, S., Titus, M.A., Soll,
D.R., and Firtel, R.A. (2002). Requirement of a vasodilator-stimulated
phosphoprotein family member for cell adhesion, the formation of
filopodia, and chemotaxis in dictyostelium. J. Biol. Chem. 277,
49877–49887.
Harris, E.S., Li, F., and Higgs, H.N. (2004). The mouse formin, FRLalpha, slows actin filament barbed end elongation, competes with
capping protein, accelerates polymerization from monomers, and
severs filaments. J. Biol. Chem. 19, 20076–20087.
Hopmann, R., Cooper, J.A., and Miller, K.G. (1996). Actin organization, bristle morphology, and viability are affected by actin capping
protein mutations in Drosophila. J. Cell Biol. 133, 1293–1305.
Hug, C., Jay, P.Y., Reddy, I., McNally, J.G., Bridgman, P.C., Elson,
E.L., and Cooper, J.A. (1995). Capping protein levels influence actin
assembly and cell motility in dictyostelium. Cell 81, 591–600.
Jacinto, A., and Wolpert, L. (2001). Filopodia. Curr. Biol. 11, R634.
Jung, G., Remmert, K., Wu, X., Volosky, J.M., and Hammer, J.A.,
3rd. (2001). The Dictyostelium CARMIL protein links capping protein
and the Arp2/3 complex to type I myosins through their SH3 domains. J. Cell Biol. 153, 1479–1497.
Katoh, K., Hammar, K., Smith, P.J., and Oldenbourg, R. (1999). Birefringence imaging directly reveals architectural dynamics of filamentous actin in living growth cones. Mol. Biol. Cell 10, 197–210.
Kojima, S., Vignjevic, D., and Borisy, G.G. (2004). Improved silencing
vector co-expressing GFP and small hairpin RNA. Biotechniques
36, 74–79.
Kureishy, N., Sapountzi, V., Prag, S., Anilkumar, N., and Adams,
J.C. (2002). Fascins, and their roles in cell structure and function.
Bioessays 24, 350–361.
Kwiatkowski, D.J. (1999). Functions of gelsolin: motility, signaling,
apoptosis, cancer. Curr. Opin. Cell Biol. 11, 103–108.
Pruyne, D., Evangelista, M., Yang, C., Bi, E., Zigmond, S., Bretscher,
A., and Boone, C. (2002). Role of formins in actin assembly: nucleation and barbed-end association. Science 297, 612–615.
Remmert, K., Olszewski, T.E., Bowers, M.B., Dimitrova, M., Ginsburg, A., and Hammer, J.A., 3rd. (2004). CARMIL is a bona fide
capping protein interactant. J. Biol. Chem. 279, 3068–3077.
Rogers, S.L., Wiedemann, U., Stuurman, N., and Vale, R.D. (2003).
Molecular requirements for actin-based lamella formation in Drosophila S2 cells. J. Cell Biol. 162, 1079–1088.
Rorth, P. (2003). Communication by touch: role of cellular extensions
in complex animals. Cell 112, 595–598.
Rottner, K., Behrendt, B., Small, J.V., and Wehland, J. (1999). VASP
dynamics during lamellipodia protrusion. Nat. Cell Biol. 1, 321–322.
Schafer, D.A., and Cooper, J.A. (1995). Control of actin assembly at
filament ends. Annu. Rev. Cell Dev. Biol. 11, 497–518.
Schafer, D.A., Mooseker, M.S., and Cooper, J.A. (1992). Localization
of capping protein in chicken epithelial cells by immunofluorescence
and biochemical fractionation. J. Cell Biol. 118, 335–346.
Schafer, D.A., Korshunova, Y.O., Schroer, T.A., and Cooper, J.A.
(1994). Differential localization and sequence analysis of capping
protein beta-subunit isoforms of vertebrates. J. Cell Biol. 127,
453–465.
Schafer, D.A., Jennings, P.B., and Cooper, J.A. (1996). Dynamics of
capping protein and actin assembly in vitro: uncapping barbed ends
by polyphosphoinositides. J. Cell Biol. 135, 169–179.
Schafer, D.A., Welch, M.D., Machesky, L.M., Bridgman, P.C., Meyer,
S.M., and Cooper, J.A. (1998). Visualization and molecular analysis
of actin assembly in living cells. J. Cell Biol. 143, 1919–1930.
Sizonenko, G.I., Karpova, T.S., Gattermeir, D.J., and Cooper, J.A.
(1996). Mutational analysis of capping protein function in Saccharomyces cerevisiae. Mol. Biol. Cell 7, 1–15.
Strasser, G.A., Rahim, N.A., VanderWaal, K.E., Gertler, F.B., and
Lanier, L.M. (2004). Arp2/3 is a negative regulator of growth cone
translocation. Neuron 43, 81–94.
Kwiatkowski, A.V., Gertler, F.B., and Loureiro, J.J. (2003). Function
and regulation of Ena/VASP proteins. Trends Cell Biol. 13, 386–392.
Svitkina, T.M., and Borisy, G.G. (1998). Correlative light and electron
microscopy of the cytoskeleton of cultured cells. Methods Enzymol.
298, 570–592.
Lanier, L.M., Gates, M.A., Witke, W., Menzies, A.S., Wehman, A.M.,
Macklis, J.D., Kwiatkowski, D., Soriano, P., and Gertler, F.B. (1999).
Mena is required for neurulation and commissure formation. Neuron
22, 313–325.
Svitkina, T.M., and Borisy, G.G. (1999). Arp2/3 complex and actin
depolymerizing factor/cofilin in dendritic organization and treadmilling of actin filament array in lamellipodia. J. Cell Biol. 145, 1009–
1026.
Lebrand, C., Dent, E.W., Strasser, G.A., Lanier, L.M., Krause, M.,
Svitkina, T.M., Borisy, G.G., and Gertler, F.B. (2004). Critical role of
Ena/VASP proteins for filopodia formation in neurons and in function
downstream of netrin-1. Neuron 42, 37–49.
Svitkina, T.M., Bulanova, E.A., Chaga, O.Y., Vignjevic, D.M., Kojima,
S., Vasiliev, J.M., and Borisy, G.G. (2003). Mechanism of filopodia
initiation by reorganization of a dendritic network. J. Cell Biol.
160, 409–421.
Loisel, T.P., Boujemaa, R., Pantaloni, D., and Carlier, M.F. (1999).
Reconstitution of actin-based motility of Listeria and Shigella using
pure proteins. Nature 401, 613–616.
Taoka, M., Ichimura, T., Wakamiya-Tsuruta, A., Kubota, Y., Araki, T.,
Obinata, T., and Isobe, T. (2003). V-1, a protein expressed transiently
during murine cerebellar development, regulates actin polymerization via interaction with capping protein. J. Biol. Chem. 278, 5864–
5870.
Loureiro, J.J., Rubinson, D.A., Bear, J.E., Baltus, G.A., Kwiatkowski,
A.V., and Gertler, F.B. (2002). Critical roles of phosphorylation and
actin binding motifs, but not the central proline-rich region, for Ena/
vasodilator-stimulated phosphoprotein (VASP) function during cell
migration. Mol. Biol. Cell 13, 2533–2546.
Lu, M., Witke, W., Kwiatkowski, D.J., and Kosik, K.S. (1997). Delayed
retraction of filopodia in gelsolin null mice. J. Cell Biol. 138, 1279–
1287.
Mallavarapu, A., and Mitchison, T. (1999). Regulated actin cytoskeleton assembly at filopodium tips controls their extension and retraction. J. Cell Biol. 146, 1097–1106.
Moseley, J.B., Sagot, I., Manning, A.L., Xu, Y., Eck, M.J., Pellman,
D., and Goode, B.L. (2004). A conserved mechanism for Bni1- and
mDia1-induced actin assembly and dual regulation of Bni1 by Bud6
and profilin. Mol. Biol. Cell 15, 896–907.
Mullins, R.D., Heuser, J.A., and Pollard, T.D. (1998). The interaction
of Arp2/3 complex with actin: nucleation, high affinity pointed end
Vignjevic, D., Yarar, D., Welch, M.D., Peloquin, J., Svitkina, T., and
Borisy, G.G. (2003). Formation of filopodia-like bundles in vitro from
a dendritic network. J. Cell Biol. 160, 951–962.
von Bulow, M., Rackwitz, H.R., Zimbelmann, R., and Franke, W.W.
(1997). CP beta3, a novel isoform of an actin-binding protein, is a
component of the cytoskeletal calyx of the mammalian sperm head.
Exp. Cell Res. 233, 216–224.
Zigmond, S.H., Evangelista, M., Boone, C., Yang, C., Dar, A.C., Sicheri, F., Forkey, J., and Pring, M. (2003). Formin leaky cap allows
elongation in the presence of tight capping proteins. Curr. Biol.
13, 1820–1823.